The goal of this program is to improve the diagnostic sensitivity and especially the specificity of autofluorescence endoscopes by examining additional information on the molecular constituents of the tissue being observed as determined by a fiberoptic Raman probe. The combined instrument will first use autofluorescence imaging to quickly identify areas of tissue which are likely to be dysplastic or cancerous. In a standard bronchoscopy procedure a small sample of tissue from this area would be taken with forceps passed through the narrow biopsy channel of the endoscope. In this program, a fiberoptic Raman probe will first be passed through this biopsy channel and pressed lightly against the suspect area of tissue. A 1-second pulse of narrowband, 830 nm light will be directed onto the tissue through a central fiber and lens assembly at the distal tip of the probe. A long-pass filter in the tip of the probe will block the backscattered 830 nm excitation wavelength but pass the longer wavelength tissue fluorescence and Raman-shifted light into a ring of collection fibers. These collection fibers will carry the scattered light back to a spectrometer which will disperse it into a characteristic Raman "fingerprint" spectrum. This spectrum contains many individual peaks corresponding to the concentration of specific molecules in the tissue. The standard biopsy will then be taken from the tissue site and sent to a pathology lab to be examined for evidence of dysplasia or cancer. This program will catalog the measured spectra and correlate them with the results of pathology. Once the catalog of spectra contains samples of normal, dysplastic and cancerous tissue fingerprints it will be possible to begin to predict, in real-time, what the results of pathology will be. A large number of spectra will need to be collected, however, from many different stages of discovered cancers before the predicted results are likely to be reliable. The use of autofluorescence imaging during bronchoscopy procedures has been shown to improve the sensitivity of cancer diagnosis over that which can be achieved by white light observation alone. This means there are few instances where a truly cancerous or dysplastic area of tissue is seen to be normal (a false negative determination of cancer). On the other hand, the method often classifies normal tissue as cancerous (a false positive determination) which leads to an additional risk to the patient from unnecessary biopsies, wasted time during the procedure and additional costs. The additional information from the Raman probe is expected to be able to identify many of these false positive sites as truly normal, increasing the specificity of the tissue characterization. Once the method is proven to be effective it will reduce patient risk, reduce the time required for procedures and reduce the cost of unnecessary biopsies. [unreadable] [unreadable] [unreadable]